Prospects for an Improved Lense-Thirring Test with SLR and the GRACE Gravity Mission John C. Ries, Richard J. Eanes, Byron D. Tapley and Glenn E. Peterson* Center for Space Research, The University of Texas at Austin Austin, TX, USA * now at the Aerospace Corporation, Los Angeles, CA, USA Abstract The theory of General Relativity predicts several non-Newtonian effects that have been observed by experiment, but one that has not yet been directly confirmed with confidence is the existence of the Lense-Thirring precession of an orbit due to the gravitomagnetic field. Previous analyses using satellite laser ranging (SLR) data to LAGEOS-1 and LAGEOS-2 are limited by optimistic and unprovable assumptions regarding the magnitude and correlation of the errors in the low degree geopotential harmonics. Now that the joint NASA-DLR GRACE (Gravity Recovery and Climate Experiment) mission has already determined a dramatically improved geopotential model, we can examine the expected improvements in the Lense-Thirring experiment. Introduction In General Relativity, the presence of mass causes 4-dimensional space-time to curve (Figure 1). A satellite that would travel in a straight line on the flat Minkowskian space (i.e., unaccelerated) still travels a straight line on the curved manifold but only in the local sense. To an external observer, the path of the satellite looks curved and is interpreted as orbital motion. Figure 1. Space in the presence of an object with mass. However, when the influencing body is rotating, the local inertial frame of the satellite can be thought of as being “dragged” in the direction of rotation (Figure 2). For this reason, the effect of the central body’s rotation on a satellite’s orbit is often called “frame-dragging” [Lense and Thirring, 1918]. It has also been called gravitomagnetism through a different line of reasoning [Ciufolini, 1986]. Just as a spinning charge produces a magnetic field, a spinning mass (angular momentum) causes the creation of mass currents and hence the production of a gravitomagnetic field (Figure 3). 13th International Workshop on Laser Ranging: Proceedings from the Science Session 1 Figure 2. Influence of rotation on the background space-time. È Ê M ¥ r ˆ ˘ È Ê G J ¥ r ˆ ˘ F = q v ¥ — ¥ Á F = m v ¥ — ¥Á (-(1+ g) cb B 0 Í Ë r 3 ¯ ˙ LT 0 Í Ë c2 r 3 ¯ ˙ Î ˚ Î ˚ Spinning Electron Spinning Mass Figure 3. Similarity between electromagnetic and gravitomagnetic fields (M is the magnetic moment and Jcb is the angular momentum of the central body, q0 and m0 are the charge and rest mass, r and v are the position and velocity, G and c are the gravitational constant and speed of light, g is a relativity parameter). Observing the effect of the gravitomagnetic field is important for several reasons. First, this field can be considered as a new field of nature, which is analogous to the magnetic field in electrodynamics. Second, the measurement of the gravitomagnetic field will provide experimental support for the general relativistic formulation of the Mach principle that the local inertial frames are determined or at least influenced by the mass-energy distribution and currents in the universe [Wheeler, 1988]. Finally, a demonstration of this effect would be of significant importance for high-energy astrophysics. Some theories of energy storage, power generation, jet formation and jet alignment of quasars and active galactic nuclei are based on the existence of the gravitomagnetic field of a supermassive black hole [Thorne et al., 1986] Testing for Gravitomagnetism Gravitomagnetism has two observable effects: the Schiff precession and the Lense-Thirring precession. Both can be described as the “dragging of inertial frames” but have different consequences. The Schiff effect arises from the spin-spin interaction between the satellite’s intrinsic angular momentum (i.e., its spin) and the angular 2 13th International Workshop on Laser Ranging: Proceedings from the Science Session momentum of the central body (central body spin). Both spinning bodies are producing a gravitomagnetic field but the larger central body’s field is forcing a change in the local inertial frame which can be observed through monitoring the change in the smaller body’s field. Equivalently, this means that the Schiff effect causes a precession in the spin axis of an orbiting satellite (or gyroscope). The NASA Relativity Experiment, Gravity Probe-B [Everitt et al., 1980], is designed to measure this effect through the monitoring of the precession of onboard ultra-precise gyros to better than 1% (Figure 4). Figure 4. The GP-B relativity experiment by Stanford University, planned for launch in 2003, is expected to verify both geodesic precession and frame dragging to much higher precision than currently possible. The Lense-Thirring effect on the other hand, is generated solely by the angular momentum of the central body. As the Earth spins, its “drag” on space-time produces a systematic change in the line of nodes and in the perigee, a perturbation that is equivalent in form to a zonal geopotential effect. Direct measurement of the Lense-Thirring precession is a difficult proposition. The effect is extremely small, ~31 mas/yr for the LAser GEOdynamics Satellite (LAGEOS) node [Ciufolini, 1986]. The dominant source of difficulty in detecting this small signal lies with the errors in the even zonal portions of the geopotential. The even zonals contribute a secular motion to the satellite’s orbit which, if perfectly known, could be removed from the orbit analysis. However, the even zonals are not known exactly; the current error in the knowledge of J2 is on the same order of magnitude as the Lense-Thirring effect. In addition to the static error portion of the gravity field, errors in the time-varying portions of the even zonals (secular, tidal and seasonal variations) also swamp the precession due to their magnitude relative to the Lense-Thirring effect. In order to overcome these deficiencies, Ciufolini [1986] proposed using two LAGEOS type satellites in supplementary orbits to measure the Lense-Thirring precession where a proposed LAGEOS-3 satellite would be placed into an orbit identical with that of LAGEOS-1 but with a supplementary inclination (Figure 5). Placing the two laser-ranged satellites in this configuration provides an exact cancellation in the precession due to the zonal coefficients. The largest remaining error source, tidal effects that do not cancel, would then average out over time. Ries [1989] examined the LAGEOS-1/3 configuration and obtained a level of error in the Lense-Thirring recovery of ~8%. Peterson [1997a] revisited the analysis using more up-to-date models and estimated that an uncertainty of ~4% could be achieved. Unfortunately, it does not appear likely that a third LAGEOS satellite will be launched in the near future to support this experiment. Consequently, any analysis must be done using SLR tracking of existing satellites. 13th International Workshop on Laser Ranging: Proceedings from the Science Session 3 Figure 5. LAGEOS-1/LAGEOS-3 supplementary orbit configuration. Recent Analysis Recently, Ciufolini et al. [1998] asserted a confirmation of the Lense-Thirring effect at the 20% level using the EGM96 gravity model [Lemoine et al., 1998] and SLR tracking to LAGEOS-1 and LAGEOS-2. Their method used three pieces of information (LAGEOS-1 node rate, LAGEOS-2 node rate and LAGEOS-2 perigee rate) to determine the LT effect along with J2 and J4. As a result, it was only necessary to consider the errors in the EGM96 gravity model for J6, J8…. This idea was innovative, but there are significant uncertainties in the assumptions used for assessing the experiment error, especially regarding the contribution of the zonals [Ries et al., 1998]. One problem is the use of the LAGEOS-2 perigee to eliminate the errors in J4. This introduces the effect of a number of non- gravitational forces for which the models are uncertain, making the subsequent error analysis difficult to prove. A more serious problem is the use of a very favorable negative correlation between zonals in EGM96 (the result of poor separation of the zonals in the gravity solution) to reduce the error introduced by the gravity model from approximately 50% to 13%. The EGM96 covariance, like any gravity solution covariance, is only an approximate estimate of the errors in the gravity solution; it cannot be considered to be an exact representation of the magnitude or correlation of the error in the individual coefficients. Further, there is no reason to expect that the errors in the EGM96 gravity model (which is a multi-decade mean gravity solution) are representative of the actual errors in the gravity model during the period of the Lense-Thirring analysis, in light of known secular, seasonal and decadal variations in the Earth’s gravity field. A more realistic error assessment would not rely on the cancellation of the errors due to a fortunate correlation, and it probably would treat the magnitude of the errors in the higher degree zonals given by the EGM96 covariance with some caution. This would lead to an estimated error in the current determination of the Lense-Thirring precession of at least 50 to 100%, if not larger. While there may be various views about this, any experiment based on current gravity models will be plagued with uncertainties regarding the error analysis. A much more solid foundation is necessary for a confident 4 13th International Workshop on Laser Ranging: Proceedings from the Science Session verification of General Relativity. What is needed is a gravity model so accurate that contributions from errors in the zonals are not significant. This removes the need to use the LAGEOS-2 perigee rate, resulting in a much cleaner signal from just the node rates.